[CANCER RESEARCH 44, 3961-3969, September 1984]

Emergence of Permanently Differentiated Cell Clones in a Human ColonieCancer Cell Line in Culture after Treatment with Sodium Butyrate1

Chantai Augeron and Christian L. Laboisse2

Laboratoire de Biologie et de Physiologie des Cellules Digestives ft/239 Institut National de la Santéet de la Recherche Médicale),Facultéde MédecineXavier Bichai,16, Rue Henri Huchard, 75018 Paris, France

ABSTRACT

The human colonie cancer cell line HT29 is ^differentiated instandard culture conditions (Dulbecco's medium: 10% fetal bo

vine serum). These cells were cultured in 5 mw sodium butyratefor 9 days; then they were trypsinized and subcultured in sodiumbutyrate for an additional 14 days. Multinucleation occurredduring this second phase of the treatment. The cells were thentransferred to standard medium and multinucleation disappeared. Morphological changes appeared 10 to 12 days afterreturn to standard culture conditions; some cells flattened andbecame more adherent to the bottom of the flasks. These alteredcells divided actively and formed "flat foci" interspersed among

the densely packed undifferentiated HT29 cells. This alteredphenotype persists after more than 24 months of culture instandard medium. Clonal cell lines were established from theseflat foci-forming cells and characterized. These clonal lines exhib

ited morphological cell polarity defined by an apical cell surfaceseparated by junctional complexes from the basolateral cellsurface. Functional differentiation did also occur since someclonal lines formed domes representing active transepithelialtransport, and others exhibited massive mucus secretion. Inconclusion, our findings indicate that permanently differentiatedcell populations emerged in a colonie cancer cell line after sodiumbutyrate treatment. These new clonal lines will be useful in futuremodels for the study of differentiation programs of both normaland cancerous colonie cells.

INTRODUCTION

Cancer cells are often characterized by a block of differentiation which can be overcome in some cases by differentiationinducers. Cancer cell lines of somatic origin are useful modelsfor the study of conditions allowing full expression of differentiation characters. These models fall into 3 broad classes: (a) cancercells responsive to naturally occurring substances which arenecessary for the maintenance of the differentiated state of theirnormal counterparts, e.g., PC12 pheochromocytoma cells whosedifferentiation into neuronal-like cells is dependent upon the

presence of nerve growth factor in the medium (8); (o) cell lineswhich are committed to differentiate by short-term treatmentwith various chemicals such as polar compounds (dimethyl sulf-oxide), NaBT,3 or retinoic acid, e.g., Friend cells (7) or HL 60

1Supported by Institut National de la Santéet de la Recherche Médicale(CRL

817022), théAssociation Charles Debray, and grants from the Association pour leDéveloppement de la Recherche de la Cancer, théFacultéXavier Bichat, and théLigue Nationale Françaisecontre le Cancer.

2To whom requests for reprints should be addressed.3The abbreviations used are: NaBT, sodium butyrate; FBS, fetal bovine serum;

SM, standard medium; STV, isotonically buffered saline, trypsin, Versene; DME,Dulbecco's modified Eagle's medium.

Received February 6, 1984; accepted June 4, 1984.

cells (4); and (c) cancer cells whose differentiation is induced bychanging nutritional conditions, e.g., reversible enterocytic differentiation of colonie carcinoma cells in a medium containinggalactose (13).

In all of these models, the action of the various inducers isreversible (7,14,15). When a terminally differentiated phenotypeis induced in a culture, the transfer of the cells into an inducer-free medium allows the reconstitution of the population of self-

renewing cells from the uninduced cells (7,15). This reconstitution is accompanied by the death of the differentiated postrepli-

cative cells. Moreover, when rechallenged with the inducers,these self-renewing cells express unchanged sensitivity to the

inducer (7,15). This indicates that the cells are not permanentlyaffected by these treatments (7). However, preliminary studiesin our laboratory indicated that following a treatment with NaBTmorphologically altered cell populations emerged in a humancolonie cancer cell line. We report here that these altered cellsrepresent in fact differentiated cell populations. We show alsothat these differentiated phenotypes are permanent and stablein SM. This type of manipulation offers a promising way forestablishing clonal cell lines expressing a highly differentiatedphenotype in standard culture conditions.

MATERIALS AND METHODS

Cell Cultures

The human colonie adenocarcinoma cell line HT29 was obtained fromDr. J. Fogh and was used between passages 148 and 165 (6). HT29cells were routinely cultured in DME (Grand Island Biological Co., GrandIsland, NY) supplemented with 10% heat-inactivated FBS (Grand Island

Biological Co.). This medium was designated as SM.

Butyrate Treatment

Stock cultures of HT29 colon carcinoma cells were dissociated bytreatment with STV (0.25% trypsin:0.02% EDTA) and then 2 million cells/flask were seeded in 25-sq cm plastic flasks (Nunc or Falcon) at day 0

in SM. At Day 2 (exponential phase of growth) NaBT (Merck, Darmstadt,Germany) was added at a concentration of 5 PIM (Chart 1). Then, themedium containing NaBT was changed every 2 days. The cells weredissociated with STV at Day 11, and then after a 1:3 split, the cells wereplated in 25-sq cm flasks in DME: 10% heat-inactivated FBS, containing

NaBT (5 rtiM) (Chart 1). Then the medium containing NaBT was changedevery 2 days. At Day 25, the cells were again trypsinized with STV asdescribed above and plated in fresh SM without NaBT in 25-sq cm flasks

(Chart 1). The cells were then routinely cultured in SM. This modified cellline was designated as "HT29diff."

Control Treatment. In another type of experiment, the cells wereplated in SM at day 0, and 5 RIMNaBT was added at Day 2 as describedabove. Then the medium containing NaBT was renewed every 2 daysup to days 11 or 25 without passaging the cells. At days 11 or 25, thecells were dissociated with STV and then cultured in SM.

SEPTEMBER 1984 3961

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

C. Augeron and C. L Laboisse

NaBt Treatment

Dj (Dt

DU

^+4 H-K»)

25

D25

©

Chart 1. NaBT treatment, a, first phase of treatment; b, second phase oftreatment; c, culture in SM; , culture ¡nSM; ++, culture in NaBT; *,subculture of cells; D0, plating of cells in 25-sq cm flasks in SM; D2, addition ofNaBT to the culture medium; Du, cells subcultured in the presence of NaBT; D»,end of the treatment. Cells are transferred to SM, and then they are propagatedcontinuously in SM.

Cell Counting

To perform cell counts, it was necessary to obtain a monodispersedcell suspension. HT29 was dissociated with a brief treatment of the cellswith STV. However, the clonal cell lines (see below) were insensitive tothis treatment, and a 2-step dissociation procedure was developed to

obtain a monodispersed cell suspension: (a) the cells were incubated for7 min at 37°with 1 HIM EDTA; (o) EDTA was discarded and replaced by

STV, which allowed the cells to become round and to detach themselvesfrom the bottom of the flasks. Then the suspended cells were countedin a hemocytometer.

Plating Efficiency and Mucosecretion Index Determinations

Known numbers of single cells were plated into replicate 60- or 100-mm diameter Petri dishes containing Ham's:DME (1:1, v/v) medium

supplemented with 10% heat-inactivated FBS. The dishes were incubated for colony formation at 37° in a humidified incubator with 5%

COz:95% air. By the end of 30 days, the contents of the dishes werefixed, and mucus was stained with a metachromatic staining method(see below). Then, the colonies were counted, and the plating efficiencywas determined (number of colonies per number of cells plated x 100).The ratio of the mucus-secreting colonies to the total number of coloniesis the mucosecretion index. Only the colonies exhibiting massive meta-chromasia were considered as mucus-secreting colonies.

Cloning of HT29diff. Cells

Isolation of Flat Colonies. A suspension of cells enriched with "flatfoci"-forming cells was prepared from cultures of the HT29diff. line as

follows. The medium was removed from the culture flasks, and after 2rinses with phosphate-buffered saline (in mg/liter: NaCI, 8000; KCI, 200;

Na2HPO4, 1150; KH2PO4 200, pH 7.2), the cells were incubated withSTV at 37°until many cells surrounding the flat foci were freed from the

bottom of the flasks. This first harvest was discarded, and fresh STVwas added to the flasks. The cell suspension from the second harvestwas enriched with flat foci-forming cells and was seeded at 200 cells/dish into replicate 100-mm diameter Petri dishes containing Ham's:DME

(1:1, v/v):10% heat-inactivated FBS. Then the dishes were incubated at37°for colony formation. By the end of 20 to 30 days of incubation,

examination of the dishes with an inverted microscope disclosed flatcolonies. When they were sufficiently separated to permit isolation ofindividual colonies, they were marked by encircling with a pencil. Then,one or 2 flat colonies were isolated from each dish using the followingtechnique. The dishes were rinsed several times with phosphate-buffered

saline to remove floating and loosely attached cells. Sterile cloning ringswere secured with sterile grease around each selected colony. Thegrease effectively isolated each colony, thus minimizing contaminationwith other cells. Each colony was removed from the cloning ring withSTV and seeded in a 35-mm culture dish in DME:20% heat-inactivated

FBS. When the cells reached confluency, they were subcultured in 25-

sq cm plastic flasks and were designated as clone 1, clone 2, etc.Dilution Plating Technique. Some clonal lines were cloned a second

time using a dilution plating technique. Briefly, a single-cell suspensionwas diluted appropriately in DME supplemented with 20% heat-inacti

vated FBS and insulin (10 ¿¿g/ml;Sigma Chemical Co., St. Louis, MO) toyield a final number of 30 cells/ml. This cell suspension was dispensedinto the wells of microtest plates (96-well microtest plates; Falcon,

Oxnard, CA). Each well was inoculated with 50 n\ of the cell suspension,yielding on the average 1.5 cells/well. Those wells containing only onecell as ascertained by microscopic inspection were identified with a markon the overlying lid. The plates were then incubated at 37°,and the cells

from the selected wells were subcultured into separate 35-mm culture

dishes. Each subclone was designated with a letter following the numberof the parent clone (e.g., cl.16A, cl.16B, and so on). All of these clonallines were subcultured and kept frozen in liquid nitrogen at an earlypassage.

Mycoplasma Detection

The HT29 cells, the HT29diff. cells, and clonal lines cl.12B, cl.13G,CI.16E, CI.18A, and cl.19A were periodically tested for Mycoplasmacontamination using the DMA fluorochrome Hoechst 33258 stainingmethod (3). These cells were always found to be free of Mycoplasmacontamination.

Assessment of Differentiation Markers in Culture

Mucus Secretion. Screening of clonal lines for mucus secretion wasperformed by in situ metachromatic staining of the monolayers as follows.The culture flasks were rinsed twice with phosphate-buffered saline and

fixed in 2 changes of a modified Camoy fixative (acetic acid:absolutemethanol, 1:3) at 30 min each. Then the fixative was discarded, and theflasks were dried overnight in an oven at 60°.The metachromatic staining

of mucus was applied to the fixed monolayers according to the methodof Hess and Hollander (9) with slight modifications. Then the flasks wereair dried and observed with an inverted microscope.

The percentage of mucus-secreting cells in a monolayer was assessed

on cytocentrifuge preparations stained with the metachromatic method.Epon-embedded Monolayers. Cell monolayers were fixed in 3%

glutaraldehyde and were embedded in situ according to the method ofBrinkley and Chang (2). Semithin (2 prn) and thin (600 A) sections werecut tangential or perpendicular to the plane of culture with an LKBultrotome (LKB, Bromma, Sweden). Semithin sections were stained withtoluidine blue and Alcian blue at pH 2.5 and 1. Thin sections were stainedwith uranyl acetate and lead citrate and examined with a Zeiss EM 109electron microscope.

Dome Formation. Domes were readily observed by inspecting cultureflasks with the naked eye or with the inverted microscope.

RESULTS

Morphology of HT29 Cells in Standard Culture Conditions

When cultured in SM, HT29 cells grew as tightly packed cellswithout signs of differentiation and without dome formation (Fig.14). Staining of the culture flasks with the in situ metachromaticmethod did not reveal any mucus secretion. Quantification ofmucus secretion on cytocentrifuge slides revealed that less than0.08% of the cells were secreting (Table 1).

Morphological examination of vertical sections of Epon-embedded cultures disclosed multilayers of densely packed cellswithout polarization (Fig. 1B). At the ultrastructural level, the cellswere quite undifferentiated (Fig. 1C). They were linked by numerous desmosomes; tight junctions were never observed (Fig.10).

3962 CANCER RESEARCH VOL. 44

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table 1Quantification of mucus secretion in HT-29 cells in standard culture conditions

Percentage of mucus-secreting cells was determined on metachromaticallystained cytocentrifuge preparations. For PEa and Ml determinations, cells were

seeded in 100-rnm diameter Retri dishes, then fixed at Day 29, and submitted tothe metachromatic staining.

No. of mucus-secreting cells

PEMl

2/2630 (O.OSf75 ±7%c

0/1511 (0)* PE, plating efficiency; Ml, mucosecretion index [(number of metachromatic

colonies/total number of colonies) x 100).'' Numbers in parentheses, percentage.c Mean of 10 dishes ±S.D.

Table 2Effects of NaBT treatment on PE* and Ml of HT-29 cells

Following the NaBT treatment, the cells were plated in 60-mm diameter culturedishes in Ham:DME:10% heat-inactivated FBS. After a 22-day incubation, the cellswere fixed and metachromatically stained. Colonies were scored for PE and Mldeterminations.

No. of cells plated PE(%) Ml

1000500200

10.6"9.5 ±1.3o9.6 ±3.7"

6/212(2.8)°

6/190(3.2)0/77(0)

" PE, plating efficiency; Ml, mucosecretion index.6 Mean of 2 dishes.c Numbers in parentheses, percentage." Mean ±S.D. of 4 dishes.

Effects of NaBT on HT29 Cells

During Treatment. NaBT treatment was initiated during theexponential phase of growth (Chart 1). During this first step ofthe treatment no major morphological alteration of the cellsoccurred. At this concentration of NaBT (5 ITÃŒM),the growth ratewas reduced, and when the cells reached confluency (Day 11)they were passaged in the presence of NaBT (Chart 1). Majormorphological alterations occurred during this second phase oftreatment; massive multinucleation developed 6 days after seeding in NaBT (Day 17) as observed with the inverted microscope(Fig. 2A). At Day 21, the bottom of the flasks was covered byfoci of giant multinucleated cells, and the monolayer began todetach from the bottom. Then the cells were transferred at Day25 to new flasks in SM.

After Treatment. Following return to standard culture conditions, multinucleation disappeared and regrowth occurred asmononuclear cells. When plated at clonal density in SM, the cellsexhibited a low plating efficiency concomitant with high microse-cretion index (Table 2). When routinely cultured in 25-sq cm

flasks in SM, the cells displayed a growth pattern which wasclearly different from that of untreated HT29 cells; 10 to 12 daysafter seeding in SM, foci of whorting cells developed in theculture. These foci were flat and were interspersed amongdensely packed cells (Fig. 2,8 and C). These "foci-forming" cells

had a permanent ability to forni flat foci at confluency in standardculture conditions. These foci were readily observed with eitherthe naked eye or the inverted microscope.

Control Treatment. When HT29 cells were cultured up to 9or 23 days in NaBT without passaging the cells during thetreatment, no multinucleation occurred. After this treatment, wenever observed the emergence of flat foci. Electron microscopydid not reveal either polarized or mucus-secreting cells.

Characterization of HT29diff. Cell Line

Since the above-described changes were permanent in SM,

we designated this new cell line as the HT29diff. cell line. This

Emergence of Differentiated Clones in a Cancer Cell Line

new cell line has been in culture now for 2 years and has theability to form flat foci after confluency has occurred. The growthcharacteristics of the HT29diff. cell line are shown in Chart 2.The doubling time for HT29diff. cells was 36.21 ±7.27 hr, whilethe doubling-time value for HT29 was 41.04 ±7.51 hr. At the

plateau phase, numerous groups of flat cells developed in theculture. These flat cells appeared in circular foci, suggesting thatthey were of clonal origin. When metachromatic staining wasperformed with the HT29diff. cell line, some of these foci exhibited a strong reactivity. The occurrence of mucosecretion insome of these foci was confirmed by electron microscopy.Moreover, it was found that the cells forming these flat foci werepolarized.

Based on the strong adherence of these foci to the plasticflasks, a procedure was developed to enrich the HT29diff. cellswith flat foci-forming cells (see "Materials and Methods"). When

seeded at clonal density in culture dishes, HT29diff. cells gaverise to 2 types of colonies which were readily discernible withthe inverted microscope, "rough" colonies formed by piled up

cells forming multilayers which were morphologically analogousto those formed by untreated HT29 cells and flat colonies analogous to the flat foci.

It must be noted that the mucus-secreting colonies were flat

colonies. A typical metachromatic flat colony is shown in Fig. 2D.

Characterization of Differentiated Clonal Lines

To study their differentiation properties more extensively,these flat colonies were isolated and expansion of clonal derivatives was performed. These cloning experiments yielded 13clones. Two successive clonings were necessary to ensure theunicellular origin of these clonal lines. Therefore, 5 of these 13clones were recloned, and each yielded several subclones. Thenone subclone from each parent clone was chosen for furtheranalysis. These clonal derivatives were cultured in SM. Phasecontrast microscopy of the cultures disclosed at confluency"spindle-shaped" cells which were arranged in whorls. Five to 10

days after confluency, signs of differentiation were discerniblewith the inverted microscope. In some lines, a thin film of mucuswas excreted at the free surface of the cells (clone 16E), whilein other lines numerous domes formed (Fig. 3A), representingtransepithelial movements of solutes (clone 19A). Sometimes,

^ 500

iV 200¿S^

S 100

50

i 20

LU

oHT29HT29diff

02 4 6 810121416182022242628

DAYS IN CULTURE

Chart 2. Growth curves of HT29 and HT29diff. cells. Points, mean ±S.E. oftriplicate counts.

SEPTEMBER 1984 3963

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

C. Augeron and C. L. Laboisse

Table 3

Differentiation markers of the donai cell lines

Each cell line was cultured in DME:10% FBS. Differentiation markers wereassessed on postconfluent cell cultures. Mucus-secreting cells were scored onmetachromatically stained cytocentnfuge preparations. Cell polarization was assessed by electron microscopy. Dome formation was assessed by examiningmonolayers with the inverted microscope.

CelllineHT29

(control)Clone 12BClone 13GClone 16EClone 18AClone 19ANo.

of mucus-secretingcells2/2630

(0.08)"

123/1013 (12.14)163/1136 (14.35)501/1039 (48.22)148/1738 (8.52)

8/1294 (0.62)Polarized

cells withapical junctional

complexes+

+++

+Dome

formation+

++

•f

' Numbers in parentheses, percentage.

dome formation was associated with the presence of a mucusfilm, covering the top of the monolayer (clone 13G). As shown inTable 3, the percentage of mucus-secreting cells in the clone

16E line was 600 times higher than that of HT29 cells. Electronmicroscopic examination of the clonal lines disclosed polarizedcolumnar cells (Fig. 30), exhibiting typical junctional complexesat their apical poles. In addition to the tight junctions whichformed an attachment belt around the luminal surfaces of thecells, these junctional complexes included a row of desmosomes(Fig. 3C). These polarized cells were sometimes arranged in atypical monolayer; this type of organization of cultured cells wasreadily observed in the cell lines forming domes. Occasionally,confluent cultures of the mucus-secreting clones grew in multi

layers; one or several layers of undifferentiated basal cells werecovered by a superficial layer of columnar cells with a high degreeof polarization (Fig. 4A). These outer cells were copiously mucussecreting (Fig. 4C) and were morphologically analogous to thegoblet cells of the intestine. High-power examination of themucus-secreting cells disclosed numerous mucus vacuoles ofvarying electron density filling the cytoplasm (Fig. 4B). Histo-chemical analysis of mucus was performed on Epon-embedded

cultures of clone 16E. Histochemistry on semithin sections disclosed reactivity of the mucus with Alcian blue at pH 2.5 and notat pH 1.

These differentiated characteristics were found to be permanent in these clonal lines over a culture period of 1 year.

DISCUSSION

NaBT, a 4-carbon fatty acid, has been shown to produce

reversible changes in morphology, growth rate, and enzymeactivities of several cell types in culture (14). The responses ofseveral human colonie cancer cell lines to the addition of NaBThave been investigated (5, 10-12, 16). This compound de

creased the growth rate of colonie and rectal cancer cells inculture (lines SW 480, SW 620, and HRT-18) and caused mor

phological changes; i.e, the cells flattened and displayed longmembranous processes (10). All of these changes were foundto be reversible. NaBT also produced changes in the cell surfacemembrane protein profile of HRT-18 cells (10). It caused anincrease in the alkaline phosphatase activity of HRT-18 cells (10)and an increase in the sucrase activity of SW-620 cells (10). It

was also shown by Monta et al. (12) that the effects of NaBT onalkaline phosphatase activity in HRT-18 cells were reversible.

In contrast to the above-described effects which were concom

itant to the NaBT treatment and reversible, we observed the

emergence of altered cell populations (flat foci) following treatment of the colonie HT29 cells with NaBT, and these changeswere permanent. Clearly, the occurrence of flat foci in the culturewas associated with a particular type of treatment. This treatment involved subculturing the cells in the presence of NaBT,and this step was found to be essential for multinucleation andthen, later, differentiation to occur. In fact, when this step wasomitted, even a long treatment with NaBT was unable to inducethese flat foci. The occurrence of flat foci was observed 10 to12 days after treatment with NaBT, implying that cell divisionwas necessary for the expression of the new phenotypes. Moreover, these phenotypic changes were found to be permanent inthe HT29diff. line cultured in SM and are therefore heritable.

Since we obtained morphological evidence that some flat fociwere highly differentiated, we undertook a detailed study of theorigin and of the differentiation properties of these foci. Each flatfocus was thought to originate from a single cell because theseflat foci appeared as circular zones growing excentrically. Theexistence of flat foci-forming cells was therefore postulated, each

of them being able to generate an isolated flat focus when theyare plated at a sufficiently low density to form separated colonies.The observation of typical "flat colonies," morphologically analo

gous to the flat foci after plating HT29diff. cells at clonal density,demonstrated that our hypothesis was true. Moreover, thesecloning experiments showed that certain colonies were analogous to those formed by cloned untreated HT29 cells. Thisindicates that only a fraction of the HT29 cells was converted byour NaBT treatment. A more extensive study of the differentiationproperties of these flat foci-forming cells was performed by

isolating and subculturing several of these flat colonies. Thus,we were able to demonstrate that several clonal cell lines originating from flat foci-forming cells exhibited morphological cell

polarity defined by an apical cell surface separated by junctionalcomplexes from the basolateral cell surface. Functional differentiation also occurred since some clonal lines formed domesrepresenting active transepithelial transport and others wereactively mucosecreting. Quantification of the mucus secretionshowed that up to 48% of the cultured cells at the plateau phasewere mucus secreting in a clonal line (clone 16E). Whether thesemucus-secreting cells represent a postreplicative population arising from the maturation of an undifferentiated "stem-ceir pool

as in the normal intestine is not known yet. Experiments are inprogress to clarify this point.

These markers of functional differentiation were not only foundin the 5 fully characterized clonal derivatives but also in the otherclonal lines established from flat foci-forming cells. However, the

intensity of functional differentiation varied from clone to clone.These findings indicate the existence of interclonal differences inthe degree of expression of the potential for differentiation. Thismorphological and functional differentiation has been maintainedin the successive subcultures of the cloned lines and is thusconsidered as permanent.

The action of NaBT on HT29 cells can be explained by eithera selection mechanism or by the creation of permanent modifications in the cells.

It may be that NaBT treatment results in a selection of apopulation of cells which are spontaneously able to differentiate.A selection process would imply that the original HT29 is heterogeneous and that some cells have the genetically determinedproperty for spontaneously achieving a high level of differentiation. In fact, we observed some rare mucus-secreting cells

3964 CANCER RESEARCH VOL. 44

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

(0.08%) in the HT29 line cultured in SM. These cells, however,were distributed randomly and never formed highly differentiatedflat foci in the culture. Since this property of spontaneous differentiation would be dependent on heritable factors, one wouldexpect that cloning of HT29 cells would lead to the emergenceof a majority of undifferentiated colonies and a minority of highlydifferentiated colonies. We tested this hypothesis by analyzingthe occurence of mucus-secreting cells in 24,000 colonies origi

nating from the original HT29 cells plated at clonal density (datanot shown). Some randomly distributed mucus-secreting cells

were observed in very few colonies, but no typical metachromaticcolony was found in these experiments as would be expected ifgenetic heterogeneity were the explanation. We would suggestrather that a stochastic model like that proposed previously byTarella ef al. (15) for HL-60 cells underlies the spontaneous

differentiation in the untreated HT29. According to this model,spontaneous differentiation would occur randomly in the HT29cell line with a particular probability.

The effects of NaBT on HT29 cells may be due to the creationof permanent changes in the cells, leading to the full expressionof a differentiation program which was present otherwise butunexpressed in these cells. This second hypothesis is moreprobable for 2 reasons: (a) at the cellular level, the massivemultinucleation which preceded the occurrence of differentiationrepresents profound cellular rearrangements, and these rearrangements seem to be an obligatory step for differentiation tooccur; (fa) at the molecular level, NaBT is known to inducehyperacetylation of histones, mainly by inhibiting histone deace-

tylase (1), and it has been suggested that this process is implicated in the control of gene expression. Taken together, thesedata support the hypothesis that profound cell changes areinvolved in the effects of our NaBT treatment.

Finally, one can also hypothesize that this NaBT treatmentacts by inducing permanent modifications in a cell populationwhich would be predisposed to differentiate. To answer thisimportant question unambiguously it will be necessary to treatcloned HT29 cells with NaBT.

In conclusion, our results clearly indicate that permanentlydifferentiated cell populations emerged in a human colonie cancercell line after proper NaBT treatment. These new clonal cell lineswill be useful models in the future to study the differentiationprograms of both normal and cancerous colonie cells. In addition,comparative analysis of the clonal cell lines, differing in theirability to secrete mucus, should provide clues to the mechanismsinvolved in the regulation of mucus secretion.

Emergence of Differentiated Clones in a Cancer Cell Line

ACKNOWLEDGMENTS

We are grateful to Dr. J. Fogh for providing the HT29 cell line. We thank R.Marcassin for excellent technical assistance. We would like to thank A. Vivaldi formusical support.

REFERENCES

1. Boffa, L. C., VkJali, G., Mann, R. S., and Allfrey, V. G. Suppression of histonedeacetylation in vivo and in vitro by sodium butyrate. J. Biol. Chem., 253:3364-3366, 1978.

2. Brinktey, B. R., and Chang, J. P. Embedding in situ. In: P. F. Kruse, Jr., andM. K. Patterson, Jr. (eds.), Tissue Culture. Methods and Applications, pp. 438-443. New York: Academic Press, Inc., 1973.

3. Chen, T. R. In situ detection of mycoplasma contamination in cell cultures byfluorescent Hoechst 33258 stain. Exp. Cell Res., 104: 255-262,1977.

4. Collins, S. J., Ruscelli, F. W., Gallagher, R. E., and Gallo, R. C. Terminaldifferentiation of human promyelocytic leukemia cells induced by dimethylsulfoxide and other polar compounds. Proc. Nati. Acad. Sci. USA, 75: 2458-2462,1978.

5. Dexter, D. L., Crabtree, G. W., Stoeckler, J. D., Bavarese, T. M., Ghoda, L.Y., Rogler-Brown, T. L., Parks, R. E., Jr., and Calabresi, P. N,A/-Dimethylfor-mamide and sodium butyrate modulation of the activities of punne-metabolizmgenzymes in cultured human colon carcinoma cells. Cancer Res., 41:808-812,1981.

6. Fogh, J., and Trempe, G. New human tumor cell lines. In: J. Fogh (ed.), HumanTumor Cells in Vitro, pp. 115-141. New York: Plenum Press, 1975.

7. Friend, C., Scher, W., Holland, J. G., and Sato, T. Hemoglobin synthesis inmurine virus-induced leukemic cells in vitro: stimulation of erythroid differentiation by dimethyl sulfoxide. Proc. Nati. Acad. Sci. USA, 48: 378-382,1971.

8. Greene, L. A., and Tischter, A. S. Establishment of a noradrenergic clonal lineof rat adrenal pheochromocytoma cells which respond to nerve growth factor.Proc. Nati. Acad. Sci. USA, 73: 2424-2428, 1976.

9. Hess, M., and Hollander, F. Permanent metachromatic staining of mucus intissue sections and smears. J. Lab. Clin. Med., 32: 905-909,1947.

10. Kim, Y. S., Tsao, D., Monta, A., and Bella, A., Jr. Effect of sodium butyrate onthree human cotorectal adenocardnoma cell lines in culture. In: R. A. Malt(ed.), Cetonie Carcinogenesis, Falk Symposium 31, pp. 317-325. Lancaster,England: MTP Press, Ltd., 1982.

11. Kim, Y. S., Tsao, D., Siddiqui, B., Whitehead, J. S., Amstein, P., Bennett, J.,and Hicks, J. Effects of sodium butyrate and dimethyl sulfoxide on biochemicalproperties on human colon cancer cells. Cancer (Phila.), 45:1185-1192,1980.

12. Merita. A., Tsao, D., and Kim, Y. S. Effect of sodium butyrate on alkalinephosphatase in HRT-18, a human rectal cancer cell line. Cancer Res., 42:4540-4545,1982.

13. Pinto, M., Appay, M. D., Simon-Assmann, P., Chevalier, G., Dracopoli, N.,Fogh, J., and Zweibaum, A. Enterocytic differentiation of cultured human coloncancer cells by replacement of glucose by galactose in the medium. Biol. Cell.,44:193-196,1982.

14. Prasad, K. N., and Sinha, P. K. Effect of sodium butyrate on mammalian cellsin culture: a review. In Vitro (Rockville), 72:125-132,1976.

15. Tarella, C., Ferrerò, D., Gallo, E., Pagliardi, G. L., and Ruscetti, F. W. Inductionof differentiation of HL-60 cells by dimethyl sulfoxide: evidence for a stochasticmodel not linked to the cell division cycle. Cancer Res., 42: 445-449,1982.

16. Tsao, D., Monta, A., Bella, A., Jr., Luu, P., and Kim, Y. S. Differential effectsof sodium butyrate, dimethyl sulfoxide, and retinole acid on membrane-associated antigen, enzymes, and glycoproteins of human rectal adenocarcinomacells. Cancer Res., 42:1052-1058,1982.

SEPTEMBER 1984 3965

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

C. Augeron and C. L. Laboisse

Fig. 1. Morphological characterization of HT-29 cells growing in SM A, phase contrast microscopy of confluent HT-29 cells, x 180. 8 and C, vertical sections of HT-29 cells embedded in Epon in situ; B, semithin section showing multilayered cells. Toluidme blue, x 600. C, low-magnification electron micrograph showing unpolanzedand undifferentiated cells, x 2,200. D, HT-29 cells linked by typical desmosomes. No other kind of intercellular junction is observed in these cells cultured in SM. Electronmicrograph, x 57,500.

3966 CANCER RESEARCH VOL. 44

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

^v

fíf::W$2C "<• ¿¿£*J*-> 2 D•••••few. - - . •

Fig. 2. Morphology of HT-29 cells during the NaBT treatment {A) and the HT-29diff. line cultured in SM (B, C, and D). A, phase contrast microscopy of HT-29 cells atDay 21 of the NaBT treatment showing numerous multinucleated cells, x 180. B, low-magnification phase-contrast micrograph showing a typical circular "flat focus"

surrounded by piled-up undifferentiated cells, x 90. C, high-magnification phase-contrast microscopy showing the structure of a flat focus; spindle-shaped cells areclosely apposed. They form a monolayer which is well demarcated from the multilayered undifferentiated cells surrounding the flat focus, x 180. D, flat colony stainedwith the metachromatic method 30 days after cloning HT-29diff. cells. Dark areas, mucus-secreting cells, x 40.

SEPTEMBER 1984 3967

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

C. Augeron and C. L Laboisse

3 A

TJ

D

3CFig. 3. Characterization of donai lines forming domes (d.19A and cl.13G). A, phase-contrast microscopy of 3 domes in a postconfluent monolayer of d.19A cells.

Photograph is focussed on the adherent monolayer. x 130. B, tow-magnification electron microscopy of a vertical section of a confluent monolayer of d.19A cells.Columnar cells form a typical polarized monolayer. These cells are tightly linked at their apical pole by junctional complexes, x 5,000. C, high-magnification electronmicroscopy showing an apical junctional complex. TJ. tight junction; D, desmosome. x 160,000.

3968 CANCER RESEARCH VOL. 44

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Emergence of Differentiated Clones in a Cancer Cell Line

*" y¿s¿--» •»

Fig 4 Morphology of mucus-secreting donai lines. A, low-magnification electron micrograph of a vertical section of postconfluent cl.16E cells showing polarizedmucus secretion, x 3,000. B. ultrastructure of a mucus-secreting cell (CI.13G) exhibiting the features of a typical goblet cell; the supranuclear cytoplasm is filled withmany vacuoles of variable electron density. Vertical section, x 6,400. C, electron microscopy of a section cut parallel to the superficial layer of the cultured cells (d.16E).All the superficial cells exhibit abundant mucus secretion, x 3,000.

SEPTEMBER 1984 3969

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1984;44:3961-3969. Cancer Res   Chantal Augeron and Christian L. Laboisse  Sodium ButyrateHuman Colonic Cancer Cell Line in Culture after Treatment with Emergence of Permanently Differentiated Cell Clones in a

  Updated version

  http://cancerres.aacrjournals.org/content/44/9/3961

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/44/9/3961To request permission to re-use all or part of this article, use this link

on August 5, 2020. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from